The Tint Snob Thread

[iamlucky13]

So why do lower k values usually have higher CRI? Daylight is 6500k and isn't that the best color rendering?

Daylight isn't quite constant. The sun itself is 5700K, but daylight tends to be cooler at higher latitudes due to atmospheric scattering. I'm not quite certain I understand right, but I think the 6500K standard is based on the average of the sunlit side, and the shadowed side (illuminated by the blue-hued sky) illumination - basically this makes it a good match for most daytime conditions.

Our eye's peak sensitivity (555nm) occurs at the same wavelength a 5270K blackbody would peak at. I suspect that's relevant in terms of optimal color rendering and neutrality, but I'm not certain. I do know that I personally find 6000K slightly on the cool side of neutral.

As to why it is easier for LED to hit higher CRI at lower CCT, that I do not know. Something to do with the phosphors.

The overly simplified answer is you add more phosphor to increase the amount of the colors other than blue (the color of the bare LED) in order to increase CRI. Because the phosphors are non-blue, this shifts the tint away from blue - in short, the same technique used for increasing CRI lowers the color temperature.

Looking at spectral curves, it appears to me Cree uses a phosphor mix that peaks in the yellow-orange range, and they sometimes further push the spectrum towards red by filtering out excess yellow. In contrast, Nichia emitters seem to have peaks in the green and in the orange-red range. I suspect Nichia's phosphor is less efficient at lower CRI's, but gives them a little more control over CRI vs. CCT. As a result, there are high CRI, daylight tinted Nichia emitters, as Maukka confirmed in the Jaxman E2:
http://www.candlepowerforums.com/vb...urements-(Nichia-219B-bargain-high-CRI-18650)

Yuji also makes some daylight tinted, 90+ CRI emitters, but nothing with the output we expect for flashlights. The closest product in size they offer compared to an XP emitter is only rated for about 100 lumens, and I don't think the solder pads match up with an XP footprint.

 

[archimedes]

I would be quite happy with a ~ 100 lumen light ... with fantastic tint and color rendering :shrug:

 

[Zak]

There's a version of Nichia 219B that's typically over 95 Ra and 90 R9, but the tint is distinctly rosy. I kind of like the rosy tint indoors, but it looks a bit odd to me in outdoor environments with a lot of green vegetation.

 

[staticx57]

There's a version of Nichia 219B that's typically over 95 Ra and 90 R9, but the tint is distinctly rosy. I kind of like the rosy tint indoors, but it looks a bit odd to me in outdoor environments with a lot of green vegetation.

219b sw45k R9080

 

[LeanBurn]

Check this out:

ThruNite TiS NW(12L mode), Yuiji modded penlight (~8L), Yuji modded Mini-Maglite AAA clone (~8L), Mini-Maglite AAA incan (rated 9L).

The Yuji penlight has semi tired Amaloops while the AAAmag-clone and Mini MagliteAAA have alkaline cells.:mecry:

 

[iamlucky13]

I would be quite happy with a ~ 100 lumen light ... with fantastic tint and color rendering :shrug:

I know, but those of us in this thread are not representative enough to sway the market, I don't think. With a suitable MCPCB, I suppose we might be able to mod, however.

Just for kicks, I took the spectral graph from one of Yuji's BC datasheets (5000K, 95 CRI typical), scaled it, and overlayed it on Cree's XP-G2 spectral graph (I suppose 75 CRI typical?). Compare the black line to the blue line - so much better through almost the whole spectrum, except in this measurement, Yuji's underlying blue LED seems to reach not quite as far into the violet range as Cree's.

 

[archimedes]

... Photos fixed (and removed my own off-topic tangent about this) ...

Please carry on

 

[staticx57]

Cree has seemed to make some pretty good progress with the XPG3 and newer emitters but as far as emitters we are concerned with they are still behind the likes of Nichia and Yuji

 

[kaichu dento]

The overly simplified answer is you add more phosphor to increase the amount of the colors other than blue (the color of the bare LED) in order to increase CRI. Because the phosphors are non-blue, this shifts the tint away from blue - in short, the same technique used for increasing CRI lowers the color temperature.

Very good post, but I thought I'd mention that the phosphors don't boost the non-blue colors, but rather block wavelengths not cooperative in projecting light seen to be generally of a useful nature. Hence, lower output from non-cool LEDs.

 

[iamlucky13]

Very good post, but I thought I'd mention that the phosphors don't boost the non-blue colors, but rather block wavelengths not cooperative in projecting light seen to be generally of a useful nature. Hence, lower output from non-cool LEDs.

They actually do boost the non-blue colors, which are otherwise almost completely absent. They absorb some of the blue light, and use that energy to re-emit in a range of colors. The most commonly used phosphor re-emits a light that is strongest in the yellow range, but includes other shades, too. That's part of the reason for the big valley in the graph of the Cree spectrum - the phosphor they're using emits only very weakly there.

The output is lower in large part because the luminescence that the phosphors undergo is not perfectly efficient.

That said, if you look at the below-3700K Cree spectrum in the graph I posted, you'll notice that instead of a smooth hump, it has a bit of a tilt to the right starting at 530 nm. I think Cree must also include some filtering material in their phosphor to block a small portion of the green, as you suggested, in order to avoid having warm colors achieved by a single phosphor type look too pukey. This is a technique called notch filtering, and they've used it in some of their True White household bulbs, too.

That's as opposed to the double hump seen in Yuji and Nichia spectra, which suggests they're blending multiple types of phosphor, with peaks in the green and orangish-red ranges.

 

[iamlucky13]

Cree has seemed to make some pretty good progress with the XPG3 and newer emitters but as far as emitters we are concerned with they are still behind the likes of Nichia and Yuji

I'd say the main difference is Cree is focusing on high output levels, and the others more on high color rendering.

The XP-L can handle nearly twice the power as the Nichia 219C, in the same size footprint.

I will be very interested to see if the high-CRI version of the XP-L2 that Zebralight is starting to use in their newest lights really does close the gap in the color rendering performance, or if it achieves a comparably high CRI rating but doesn't quite match the real world tint quality.

 

[kaichu dento]

They actually do boost the non-blue colors, which are otherwise almost completely absent. They absorb some of the blue light, and use that energy to re-emit in a range of colors. The most commonly used phosphor re-emits a light that is strongest in the yellow range, but includes other shades, too. That's part of the reason for the big valley in the graph of the Cree spectrum - the phosphor they're using emits only very weakly there.

The output is lower in large part because the luminescence that the phosphors undergo is not perfectly efficient.

That said, if you look at the below-3700K Cree spectrum in the graph I posted, you'll notice that instead of a smooth hump, it has a bit of a tilt to the right starting at 530 nm. I think Cree must also include some filtering material in their phosphor to block a small portion of the green, as you suggested, in order to avoid having warm colors achieved by a single phosphor type look too pukey. This is a technique called notch filtering, and they've used it in some of their True White household bulbs, too.

That's as opposed to the double hump seen in Yuji and Nichia spectra, which suggests they're blending multiple types of phosphor, with peaks in the green and orangish-red ranges.

Hmmm, lots more to learn about everyday. I'd always held the oversimplified view that blocking was going on, which I suppose would have actually resulted in virtually no light being allowed to escape.

Thanks for the layman-friendly explanation!

 

[archimedes]

I think I may have posted the following link somewhere else here recently, but it is a reasonable summary ...

https://www.digikey.com/en/articles/techzone/2011/aug/whiter-brighter-leds

 

[Zak]

Remove the phosphor from a white LED and it's pure blue. These photos show phosphor partially scraped off to reveal the underlying blue. Shine the right shade of blue or UV on a white LED and you'll see the phosphor glow yellow or orange.

 

[kaichu dento]

Remove the phosphor from a white LED and it's pure blue. These photos show phosphor partially scraped off to reveal the underlying blue. Shine the right shade of blue or UV on a white LED and you'll see the phosphor glow yellow or orange.

There's an even easier way with some of the LED light bulbs that have the phosphors in the plastic lenses.

 

[noboneshotdog]

Hey Gentleman, I purchased a R50pro with a 5000k XHP70 from Vinh a while ago. The hot spot is creamy white but the spill is a bit blue. Is blue spill typical of this led? What led would people prefer in this light? I am a hvac tech and use it at work indoors primarily.

Thanks for and answer regarding the blue spill and suggesting a good indoor floody led.

 

[Zak]

Tint shift from hotspot to spill is typical of most domed Cree LEDs, especially the XHP series. The MT-G2 is better about this, and could probably be made to work in an R50 Pro, but has lost popularity to the XHP70 because it makes more heat and less light. Shaving the dome often improves the beam characteristics of the XHP70, but not always. Vinh offers this on some lights and might do others by request. Replacing the lens with a flashlightlens.com UCL or a plain lens that doesn't have an anti-reflective coating might also improve the tint of the spill. Finally, there's the Nichia 144A, but in some reflectors, it also has bad tint shift across the beam.

 

[noboneshotdog]

Tint shift from hotspot to spill is typical of most domed Cree LEDs, especially the XHP series. The MT-G2 is better about this, and could probably be made to work in an R50 Pro, but has lost popularity to the XHP70 because it makes more heat and less light. Shaving the dome often improves the beam characteristics of the XHP70, but not always. Vinh offers this on some lights and might do others by request. Replacing the lens with a flashlightlens.com UCL or a plain lens that doesn't have an anti-reflective coating might also improve the tint of the spill. Finally, there's the Nichia 144A, but in some reflectors, it also has bad tint shift across the beam.

Wow! Thanks for the informative reply. Will have to think about this one for a bit. Maybe just buy a new light. Hahahaha.

 

[Thom2022]

I'm not a massive tint snob but I've come to like a good tint. Recently I've been experimenting with various LED's to find a good option for a build for my brother in law. The XHP50.2 80+CRI is just WOW!!!!. In daylight the hotspot is quite yellow but then spill blends seamlessly into daylight. At night it a wonderfully comfortable light to look at, very incan-esque. I was aiming it off a bridge at water 20ft below and you could see the river bed 3-5ft below the surface clear as day! Absolutely perfect detail! Well worth a punt for anyone out there who likes their lights in the warm side.

 

[jon_slider]

Re: The Tint Snob Thread .

Not All Nichias are Pink

Im all about pink tint.
I always prefer tint below the BBL, not above. This is true for me at all color temperatures: Warm White, Neutral White, and Cool White.

Left to right:

1. Neutral White 4000 kelvin color temperature, High CRI, with a Rose tint (below the BBL).
My favorite tint! This is the Nichia tint I LOVE! Not necessarily my favorite CCT, that depends on ambient light.

2. Cool White 6000k CCT, Low CRI, with a Green tint (above the BBL)
The brightest, but also lowest Red Lumen Content. This is a typical "pure white" LED (marketing spin on cool white). I find it unappealing, since I discovered Pink Nichias.

3. Neutral White 4500k, High CRI, Yellow/green tint (above the BBL)
This typical yellow/green tint makes me sad, I hate yellow/green tint (above the BBL). It is a penalty of chasing high lumens. This is a High CRI Nichia I love to hate.

Can you spot the BBL in the pic below:?
as an example, if you follow the 4000k line, notice how a 4000k LED could be either yellow or pink tinted, and how the 6000k could be either green or violet tinted.
that is due to where the LED landed when its tint bin was measured, above or below the ideal BBL (extra credit if you can explain to a layperson, what the heck a BBL is ;-))

backstory on tint bin differences regarding N219b vs N219c, also 200 Bin vs 220 Bin